Unusual Clinical Presentation of Carcinoma
927
FIG. 33.37.
Lymphatic tumor emboli in the nipple.
A,B:
Carcinoma cells in lymphatic channels in
the stroma of nipples in patients who did not have clinical manifestations of inflammatory carci-
noma. Recurrent carcinoma in such patients is likely to have inflammatory features clinically.
A
B
slightly less acute clinical course, but ultimately not a better
survival, than women with classical primary IBC.
Reports in the 1980s and early 1990s described a substan
tial improvement in prognosis when compared with earlier
data for primary IBC with 5-year survivals ranging from 25%
to 48%
273–276
and 10-year survival of 32%.
276
Certain clini
cal features of the disease (including nodal metastases and
chest wall adhesion) at presentation and stage at the time of
diagnosis have an important influence on prognosis, with a
better 5-year survival observed in those with localized than
with regional disease.
274,276
Mastectomy was shown in the past to be ineffective by itself
and was rarely performed for IBC,
234
but it is now considered
an integral part of the multimodality treatment program. Mas
tectomy has been relatively effective for obtaining local control
of the primary tumor when preceded by combination chemo
therapy followed by radiation.
250,277–281
Treatment with anthra
cycline-based neoadjuvant chemotherapy followed by surgery
and/or radiotherapy can result in local control in at least 80%
of patients and 5-year survival rates greater than 50%.
282
Radiation and chemotherapy often cause a diminution
in the clinical manifestations of IBC prior to a mastectomy.
The effects of treatment include some or all of the follow
ing: decrease or elimination of erythema, reduction in breast
size, loss of cutaneous edema, and decrease in the size of a
palpable tumor if present. Enlarged ALNs may also become
smaller. In a few patients, clinical signs disappeared en
tirely (clinical complete response), but residual carcinoma
was found microscopically at mastectomy in virtually all
cases.
250,275,278,280,281,283
Clinical complete response has been
reported in 12% to 52% of patients, with pathologic com
plete response in 4% to 33%.
282
Various sequences of treat
ment have been shown to reliably improve local control and
DFS. Amelioration of edema, erythema, breast enlargement,
and tumor mass has been associated with a relatively longer
DFS than has been achieved in patients who do not respond
to neoadjuvant treatment.
273,278,284
The clinical description of response to treatment does
not always correlate well with the pathologic findings in the
mastectomy specimen. Considerable residual tumor, often
with substantial lymphatic tumor emboli, may persist de
spite what appears to be a complete clinical response. Alter
natively, women reported to have partial or minimal clinical
response may prove to have little or no microscopically de
monstrable tumor. In the latter situation, one typically finds
substantial alterations in the mammary parenchyma where
the tumor has been destroyed (Fig. 33.38). These changes,
varying from simple fibrosis to chronic granulomatous in
flammation, are described in detail in Chapter 41. Similar
effects have been observed in ALNs that were pathologically
devoid of tumor after treatment although there appeared to
be little response clinically.
It has been observed that pathologic findings in the mas
tectomy specimen predict prognosis more accurately than
the clinical assessment of response to treatment.
284,285
In one
study, “therapeutic response parameters” associated with
the most favorable outcome were complete regression after
induction therapy within 8 months of diagnosis and com
plete regression of inflammatory symptoms within 3 months
of neoadjuvant therapy.
276
Patients who exhibit a good re
sponse clinically and pathologically appear to have the best
prognosis.
The number of involved ALNs may be a particularly
important prognostic factor, and for this reason surgi
cal staging has been advocated.
285
However, pretreatment
with chemotherapy may result in downstaging of axillary
involvement in patients who have a response that destroys
axillary metastases without leaving residual fibrosis in the
nodes. A cytokeratin immunostain can be employed to
detect microscopic residual carcinoma in ALNs or in the
